Interaction of Beta Particles with Matter - Analytical Chemistry (ACS

Determination of Micro Rhodium Film Thickness and of Gold Plating Thickness on Printed ... J. M. De Blasi , M. Delfino , D. K. Sadana , K. N. Ritz , M...
1 downloads 0 Views 841KB Size
s

= specific gravity

(3.12 for

C.P.

of bromine bromine)

Several determinations may be made without cleaning the traps. When cleaning is desired, allow the traps to thaw until the bromine is completely melted and then add lyater. Do not add water to the traps while they contain frozen bromine, because this may cause microscopic cracks to develop in the glass. Wash the contents of the traps into a beaker of water and destroy the bromine by adding ammonium hydroxide. Rub off the grease from the ball joints with a cloth and wash the traps in hot water. Dry in a n oven a t 150” C. RESULTS

Carbon Present,

Carbon Found,

%

%

0.021 0.032

0.020 0.032 0.072 0.118

0.074

The following results were obtained for carbon in two samples of C.P. bromine from different companies:

yoCarbon Company A 0.0086 0.0076 0.0083 0.0077 Av. 0.0081

To check the validity of the method further, the bromine that had accumulated in the first trap was analyzed for carbon after carefully drawing i t up with a pipet (to avoid contact with the grease). No significant amount of carbon could be detected. Varying amounts of chloroform were weighed in glass-stoppered flasks and portions of purified bromine, recovered from the traps, were added. The samples were then analyzed for carbon and the following results obtained:

Company B 0,0099 0.0100 0,0099 0.0096 0.0098

There seems t o be about 0.009% carbon present in C.P. bromine. If this carbon was present as chloroform, there would be about 0.05% chloroform in the bromine.

0.115 DISCUSSION

The dry ice is essential for the method. The use of liquid nitrogen or liquid air in place of the dry ice causes the carbon dioxide to freeze. Various salt and ice mixtures were also ineffective in freezing out the bromine. The addition of alcohol or acetone to the dry ice is not necessary, as dry ice alone in thermos beakers is equally effective. The sample must be introduced by means of a dropping device as shown in Figure 1. If the sample is passed into the combustion tube by bubbling oxygen through a small bottle containing the bromine, some oily material remains

that cannot bt: driven off without rharring. LITERATURE CITED

CAEhlICAL SOCIETY, Nashingt?? 6, D. C., “Reagent Chemicals, p. 82, 1951. Codell, M., Korwitz, G., ANAL. CHEM.28, 2006 (1956). Hillebrand, W. F., Lundell, G. E. F., Bright, H. A , , Hoffman, J. I., “Applied Inorganic Analysis,” p. 768, Wiley, New York, 1953. (4)Huntress, E. H,, “Organic Chlorine Compounds, p. 550, Wiley, New York, 1948. (5) Kirk, R. E., Othmer, D. F., “Encyc,)ppedia of Chemical Technology, Vol. 2, p. 629, Interscience, New York, Publishers, 1948. (6) Krauch, C., “Die Prufing der chemischen Reagentien auf Reinheit,” p. 66, Verlag von Julius Springer, Berlin, l89SiI ( 7 ) Merck, E., Chemical $;agents, Their Purity and Tests, p. 58, Merck & Co., Inc., New York, 1914. (8) Murray, B. L., “Standards and Tests for Reagent Chemicals,” p. 119, D. Van Nostrand, New York, 1920. (9) Pigott, E. C., “Ferrous Analysis, Modern Practice and Theory,” p. 118, Chapman & Hall, London, 1953. (10) Rosin, J., “Reagent Chemicals and Standards, p. 90, Van Nostrand. New York, 1955. (11) U. S. Steel Cor “Sampling and Analysis of C%rbon and Alloy Steels, p. 48. Reinhold, New York, 1948. RECEIVED for review September 14, 1956. Accepted December 26, 1956. AMERICAN

Interaction of Beta Particles with Matter RALPH H. MULLER University of California, 10s Alamos Scienfific laboratory, 10s Alamos, N.

,

b Of the several phenomena involved in the interaction of beta particles with matter, systematic studies on the relative backscattering of betas b y matter have shown that simple regularities obtain for all elements and compounds, and that these can be measured with high precision. The measurements have been applied to solids, liquids, and solutions. Values predicted for pure crystalline solids con b e checked to a precision close to that with which the atomic weights are known.

M.

strated by showing that the backscattering increases with increasing thickness of the target and eventually reaches a limiting value. The early studies showed a slow increase in backscattering with increasing atomic weight of the scatterer and some evidence of periodicity. There was also some evidence that the backscattering by compounds can be predicted from the intrinsic scattering of the constituent atoms (1, 3 ) . The present studies show that the backscattering of beta particles is a discontinuous function of atomic number, but strictly linear in Z within each period of the periodic system. The HEN beta particles strike matter, some of them are LLreflectedJ’ backscattering from compounds can be predicted with high precision on this and return in the general direction they basis. came from. That this behavior is not a case of ordinary reflection was demonSome general aspects of backscatter-

W

ing should be discussed before proceeding with the results of these studies. As this phenomenon deals with multiple scattering, excellent reasons could be given to show that the trajectory of a beta particle through an appreciable thickness of a substance is extremely complex, and that one could not hope to predict the over-all behavior from fundamental considerations. I n addition, the beta particles are not monoenergetic. They do not interact with matter in any single, simple manner; they are attracted and deflected in their path by the field of the nucleus and deflected by the extranuclear electrons, and those of higher energy can give rise to bremsstrahlung in the process of being slowed down. Therefore, one is generally advised to confine inquiry to the process of VOL. 29, NO. 6, JUNE 1957

e

969

single scattering, with the implication given that multiple scattering leads t o answers which are good to order-ofmagnitude reliability only. It is astonishing, then, that regularities of a relatively simple nature can be observed and reproduced with high precision. Figure 1 shows the relationship of the relative backscattering of several elements. These data had been obtained by the middle of 1953. This illustration is schematic; the actual data are given in Table I. The linear equations relating relative backscattering to atomic number within each period were sufficiently precise to establish the inflection points occurring a t atomic numbers 2, 10, 18, 36, and 54. From these values, the backscattering of many compounds was predicted and confirmed ( 5 ) . Since 1953, completely reliable reference materials have been prepared from which the relative slopes of the lines defining each period can be established with high precision.

a W

c

9 0

rn

X 0

a

m

w

-

I-

9

-I W

a

I

/

I

ATOMIC

Figure 1.

NUMBER

Relative backscattering of beta particles

Discontinuities appear precisely at atomic numbers 10, 18, 36, and 54 corresponding to rare gas configurations of neon, argon, krypton, and xenon

, SPECIMEN LUCITE B E T A SOURCE

EXPERIMENTAL

Relatively simple equipmcrit is used.

A Lucite sample stage containing a collimated source of beta particles is mounted over a methane proportional counter (Figure 2). A Mylar window about 2.0 inches in diameter admits the beta particles backscattered from the sample within the permitted geometry. The permissible geometry is poor; it represents a compromise between the desire to accommodate rather small specimens and to confine the counting primarily to those betas which are backscattered. The source consists of a truncated aluminum cylinder Kith a collimating cylindrical hole 0.5 mm. in diameter. An insert at the base provides a strontium-90 source. A thin stainless steel windoiv cuts off the 0.53-m.e.v. betas from strontium-90 and transmits primarily the 2.18-m.e.v. betas from the daughter element, yttrium-90 The methane proportional counter provides a good plateau between 3600 and 4100 volts. Conventional amplifier and binary scaler are used for manual counting. For the large number of measurements made in these studies, automatic counting and registry were eventually adopted. As shown in Figure 3, the counter could be connected to a five-decade Berkeley decimal scaler. The latter is provided with an automatic read-out which, upon receipt of a command signal, presents the count t o a 5-digit printer. For the automatic system a special timing unit was developed. This utilizes a 60-cycle Invar tuning fork and a gas-tube scaler. The scaler counts cycles from the fork; switching circuits permit the selection of any counting time from 1 to 59 minutes in 1-minute intervals and any dead time from 1 to 10 minutes. These command signals, including a reset signal, are applied automatically to the Berkeley scaler and printer. The timing is accurate to

970

ANALYTICAL CHEMISTRY

MYLPlR WINDOW

Y

TO PULSE AMPLIFIER AND SCALER

METHANE PROPORTIONAL COUNTER

Figure 2. Schematic of apparatus for measurement of beta particle backscattering

PULSE AMPLIFIER

INVAR FORK

SCALER 8 READOUT

SCALER

,?I

PRINTER

TIMING

Figure 3. particles

Counting methods for backscattering of

PULSES

beta

A . Manual operation (within dotted lines) B. Completely automatic timing and printing out

1 part in 100,000 well in excess of requirements, but inherent in the electrically driven Invar fork. The over-all precision of counting could be obtained from examination of the data. The printed data were blocked off into groups of 10 sets. The statistical error t o be expected for each set is known from the total number of counts. This applies only to the sta-

tistics of nuclear disintegration and has nothing to do with the other characteristics of the system. Statistically acceptable sets of data were then combined to give average values. I n critical comparisons of substances for which identical backscattering was expected (isomers), least squares computations of accumulated counts vs. elapsed time were made to establish

CUP FOR AQUEOUS SOLUTIONS LUCITE

REMOVABLE

CUP FOR NON-AQUEOUS SYSTEMS

~

3 0 MIL GOLD

L I Q U I D LAYER

L 2 0 MIL CELLULOSE ACETATE WINDOW

FOR BACKSCATTERING AT THICKNESS FOR TRANSMISSION (WITH PLUG1

Figure 4. Cells for backscattering and transmission: measurements of beta particles used with liquids and solutions

the average count rate. I n the early stages of the work, over-all rate was recorded b y a count-rate meter which was useful in detecting and correcting jources of count-rate drift. The intercomparison of data sets was considered to be necessary as a true criterion of over-all precision, rather than the sole criterion of precision. use of the Liquids and solutions are measured in thin-window cups. For aqueous solutions a Lucite cup with a 20-mil cellulose acetate window was used (Figure 4). For organic liquids a brass cup LTith a 5-mil beryllium windonwas used. Although the studies were concerned almost exclusively with backscattering, some absorption measurements were required, especially to distinguish between isomers. To this end, a brass plunger faced with 30 mils of gold was provided t o fit into the brass cup-beryllium window assembly. With this arrangement a thin film of liquid could be trapped between the window and the strongly reflecting gold facing. Betas traversing the liquid twice could be compared for absorption. With this method the calculation of absorption by the liquid is complicated by the need for a small backscattering correction from the thin film of liquid. An improved system has been de\-eloped ( 7 ) which consists of a vessel I\ ith two thin glass windows soldered t o the cell walls. This permits the simultaneous measurement of transmittance (absorption) and the backscattering correction. The entire systr.m is subject to many refinements and improvements, several of which are now under may. The relative backscattering 1-alues hcrein reported are characteristic of the particular experimental arrangement, particularly of the geometry of the system. It is possible to devise a system t o describe every aspect of the scattering process, including distribution angle and so forth. Such systems can provide elegant and detailed information on one or two specimens but are not particularly suitable for the examination of a large number of elements or compounds in different states of aggregation. It would be a serious defect if the information were highly dependent upon instrument parameters. Actually this

is not the case and the information can be obtained with highly diverse arrangements. One of the most useful results has been the discovery of reference materials which can be used to calibrate a given system in a short time. PRACTICAL REFERENCE POINTS

Figure 1 shows that the inflection points in the series of linear values connecting backscattering with Z occur a t atomic numbers 2, 10, 18, 36, and 54, corresponding to the rare gas configurations of helium, neon, argon, krypton, and xenon. K i t h the exception of atomic number 2, pure single crystals of sodium fluoride, potassium chloride, rubidium bromide, and cesium iodide should exhibit almost identical backscattering a t these respective inflection points. I n each case, the halogen is one atomic number below, and the alkali one atomic number above, the corresponding rare gas. Calculation and observation confirmed this completely:

Sodium fluoride Potassium chloride Rubidium bromide Cesium iodide

Z

Z, Searest Rare Gas

10 092 18 048 36 034 54 032

10 (Ne) 18 ( A ) 36 (Kr)

54(Se)

The value of Z for each crystal is computed from the intrinsic scattering of each element in the compound and its weight fraction in the compound. Single crystals of the above substances were grown in this laboratory. Precise measurements obtained with them have established the relationship of backscattering us. Z in periods 111, IV, and V, and thus have fixed the backscattering values for the 45 elements in these periods. The intrinsic values for backscattering listed in the tables in this paper are based on these cardinal reference points and are compared with direct measurements on elements as indicated. The reference points are now relied

on completely rather than individual measurements of the elements. I n few cases can the elements be obtained or fashioned as samples to the degree of purity or excellence of the single crystals. Although early measurements on the elements provided the clue to the inflection points, they are inferior in precision to the crystal values. The backscattering values listed for all the elements make no specific assumption about the state of aggregation. Thus, the values for helium, neon, argon, krypton, and xenon obviously do not refer to the gaseous state, for which the backscattering would be very small. K i t h the exception of these, for which the values would be valid in the liquid or solid state, the intrinsic scattering for each element refers to the contribution which it would make as a solid (or liquid) or in any compound of that element in proportion to its weight fraction. There is still no equally useful equivalent for the bottom of period I1 (helium) or the top of period VI (radon). The neon point is the “anchorage” for the top of period I1 and the xenon point for the bottom of period VI. I n period I1 it has been necessary to define the linear relationship by additional measurements on beryllium and carbon, and on values of oxygen. nitrogen, and fluorine derived from compounds. Thus, the accurate value for silicon, calculated from period 111, combined with precise measurements on quartz, enables one to get an accurate value for oxygen. I n similar fashion the value for fluorine can be obtained from fluorite (CaF2), and so forth. In this connection, a n almost equally reliable value for period I1 has been obtained by combining the computed values of sodium fluoride with those of a crystal of lithium fluoride. Measurements on these two crystals confirm the value. The problem in period VI is similar because no simple practical equivalent for radon ( Z = 86) has been devised. Here one still depends upon a pure metal or a compound to locate enough points to tie with the accurate xenon point. The confidence placed in period TI is enhanced. however, by a n empirical expression which relates the backscattering of the rare gas configurations and permits reasonable extrapolation to the radon point. This is not discussed in detail because there is not sufficient evidence for its general validity. However, periods I1 and VI are not defined with less certainty than periods 111, IV, and V. The direct and indirect measurements to supply the information for periods I1 and VI were even more extensive than in the other three periods. The principal advantage of the calibrating crystals VOL. 2 9 , NO. 6, JUNE 1957

971

lies in their purity and crystalline perfection and in the great ease with which they can define the backscattering for the 45 elements n-ithin their range. RELATIVE BACKSCATTERING OF ELEMENTS

Table I gives the relative backscattering in per cent for elements 2 to 83. The linear equations which apply to each period were defined by the cardinal calibration standards described above. For those elements in which a direct observation was made, the observed value is also given. For those elements which were determined indirectly -Le., in terms of a pure compoundthe source is indicated. Of the elements examined, about one third were of spectroscopic purity. Of the remainder the purity cannot be specified; they were of the highest quality obtainable. For this reason the calculated values should be emphasized because they represent the best average within each period. Experience has shown that this is justified. There is now an increasing amount of evidence in this laboratory to show that the most precise value for an element is best obtained indirectly from a pure crystalline compound containing that element. I n the values calculated for the rare earths, only one can be compared directly with a measurement on cerium. This large gap between elements 57 and 7 2 is unsatisfactory, but no other metals or crystalline compounds have been available. Several of the rare earths are now being investigated by aqueous solution techniques. The author ventures the opinion that there will be no anomaly in their behavior. No attempt has been made to examine thorium, uranium, or the transuranic elements. These are all highly radioactive and would be impossible to handle with the present technique. This does not minimize the importance or interest which attaches to these elements. Values for the rare gases, except helium, are average values. Each of the linear equations yields slightly different values; the mean is chosen and recorded because these points, marking the inflections of the discontinuous function, are probably important. I n no case do the extreme values, from which the mean is taken, differ by more than O.Olyoabsolute or a few tenths per cent relative. The relative backscattering values pertain to the geometry prevailing in these studies, but from the foregoing it can be seen that calibration for any other system is relatively easy. ORGANIC COMPOUNDS

I n Table I no value is given for hydro-

972

0

ANALYTICAL CHEMISTRY

gen, and helium has a small but definite value. Hydrogen exhibits negative backscattering, presumably due to absorption effects. Fortunately, this effect can be evaluated precisely and exact corrections can be applied to the observed backscattering. Muller (4) has described backscattering from representative organic compounds. Measurements showed that the carbon in each hydrocarbon accounts for more backscattering than is exhibited by the compound and the deficit is directly proportional to the hydrogen content. The average value so obtained provides the proper correction for other hydrogenbearing compounds. including hydrates. Deuterium exhibits a backscattering value which differs from hydrogen. At present, knowledge on this point is confined to the difference between heavy and ordinary water and between polyethylene and deuteropolyethylene,

Table I.

He Li Be B C

2 3 4 5 6 7 8

0.305 1.536 2.767 3.999 5.230 6.461 7.692

9

8.923 10.152

F Ke

10

Period 111. BS Ke Na Mg .41 Si

P S C1 A

10 11 12 13 14 15 16 17 18

10 152 11 116 12 084 13 051 14 018 14 986 15 953 16.920 17.895

Period IV. BS

-4

Xr

18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36

Kr Rb Sr Y Zr

36 37 38 39 40

K Ca

sc Ti v

Cr Mn

Fe Co Ni

cu

Zn

Ga Ge

AS

Se Br

SOLUTIONS

Extensive measurements have been made on aqueous solutions of the alkali chlorides and the sodium halides (6). Some of the general conclusions are

% Backscattering Calcd. Obsd. Period 11. BS = 1.23112 - 2.157

2

0

importance, no further discussion is contemplated until a precise yalue for the effect can be obtained. These investigations (4) have also revealed that isomers exhibit identical backscattering. This result would please most physicists and disappoint all chemists, because many isomers are easily distinguished by a dozen or more criteria. The resources of beta-particle techniques are not barren in this case, because isomers differ in the extent to which they absorb betas and apparently in direct proportion to their respective densities.

Relative Backscattering of Elements 2 to 83

Element

N

As this difference between hydrogen and deuterium is of considerable theoretical

LiF

2.770 5.223

=

7,703

Quartz, C,

8.934

CaF?, NaF

0.967312

16.OOi

+ 5.556

20.629 21.260 21.829 22.624 23.445 24,055 24.758 25,420 26.313

28,914

= 0.349882

Teflon, Be

+ 0.476

11.138 12 072 13 072 14 042

= 0.685822

17.895 18,587 19 272 19 958 20 644 21 330 22 016 22 702 23 387 24 073 24 759 25 445 26 131 26 816 27 502 28 188 28 874 29 560 30 253

Period V. BS

1.529

Source

+ 17.664

NaCl

given here. As shown in Figure 5, the relative backscattering as a function of weight fraction of solute is linear and permits extrapolation to unit weight fraction of solute, which is the value to be expected for the pure solid. The available concentrations are limited by solubility considerations. That these extrapolations are not too heroic is illustrated in Figure 5 . For the five salts which were examined, pure single crystals of three of them were available a t the time and their backscattering is indicated by the solid dots. All solutions do not exhibit linear behavior when backscattering is plotted against weight fraction, The binary system, acetonecarbon tetrachloride, has been studied over the entire concentration range ( 7 ) and, while the interpretation is not completed, there is unmistakable evidence of nonlinearity

Table 1.

Element

Xb M O

Tc Ru

Rh Pd -4g Cd In Sn Sb

Tc I

Xe

just as one finds in general for the colligative properties of this system. Although solution behavior is interesting in this case for its own sake, the examination of solutions for backscattering has been most useful primarily as a means of determining the intrinsic scattering of those compounds or elements which are difficult to obtain in large crystals. INTRINSIC

BS

z

1211

71

Pi-

Sd Pm Sm

Eu Gd Tb

QJ Ho

Tb

Hf Ta IT

70

72

73 74

Re os Ir

r-

T1

81 82

Pt Au Hg

Pb Bi Po

At

Rn

a 2 -/- b

ia

76 77 78 79 80 83 84 85 86

(1)

where BS is expressed in per cent and 2 is the atomic number. The values of the constants a and b which satisfy

yo Backscattering Calcd. Obsd.

Period I-, RS = 0.349882 + 41 32.009 42 32.359 32.709 43 44 33.059 33,409 45 33.758 46 34,107' 47 34.458 48 31.808 49 35.158 50 35.508 51 52 35.858 36.208 53 54 36.558

Er Tm

Ba La Ce

=

Relative Backscattering of Elements 2 to 83 (Confinued)

54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69

cs

OF ELEMENTS

The relative backscattering for the elements (Table I) is derived from measurement and can be expressed in each period by a linear equation of the form

1 T . 66-1 ( C o n t i n u e d )

Period I1 I11 IV V VI

z 2 t 0 10

10to18 is to 36 36to54 54 to 86

a 1.2311 0.96731 0.68582 0.34988 0.26225

b

-2.157 +0.476 +5.556 +17.664 $22.396

As described above, these values are readily obtainable for any other similar system in terms of a few cardinal calibrating crystals. These constants have been evaluated by the method of least squares from a large number of measurements on elements and compounds. The number of significant figures in the constants gives no immediate indication of their relevance. It is best inferred from their ability to predict backscattering in pure crystalline compounds. Also, the observed values given for a and b are not the best criterion of precision, because precise values, obtained from the cardinal calibrating substances, are now used. The calculated values in Table I and the above linear equations are considered to represent the most probable values for the elements. BACKSCATTERING OF C O M P O U N D S

It has been suspected for a long time 34.167 34,585 34,896 35,283 35.768 35.959

+

Xe

BACKSCATTERING

the experimental conditions in these investigations are as follows:

Peril-id VI. BS = 0.262252 22.396 36.558 36.820 37.082 37.344 37.607 3 i . 638 37.869 38.131 38.393 38.656 38,918 39.180 39 442 39.705 39.967 40,229 40.491 40.754 41.016 41.278 41.309 11 ,540 41 ,556 41.803 41 ,818 42.065 42 327 42.589 42.852 42. i90 43.114 43.083 43.376 43.638 43 531 43.901 43.962 44.163 44.148 44.425 44.687 44.950

that the backscattering of compounds is predictable from the intrinsic scattering of their constituent atoms. This investigation shows that, a t least in the case of single crystals of high purity, the backscattering of a compound can be predicted with a precision almost equal to that with which the atomic weights are known. Case I. For any compound in which all the constituent atoms fall in the same period, the backscattering is equal t o the intrinsic scattering for each element multiplied by its weight fraction in that compound and summed for all other atoms in the same way. Thus, for sodium chloride, 22.997 Bxa 35.457 Bci BS = 58.454

+

where B N and ~ BCI are the intrinsic scattering of sodium and chlorine, respectively. I n this equation 22.997 (atomic weight of sodium) divided by 58.454 is the weight fraction of sodium; similarly, 35.457/58.454 is the weight fraction of chlorine. Case 11. This simple principle does not hold for compounds in which the constituent atoms lie in different periods, a situation which may be apparent from the discontinuous nature of the backscattering us. 2 relationship (Figure 1). A more general relationship, which reduces exactly to Case I where all elements are in the same period, has been described previously ( 6 ) . An effective atomic number 2 VOL. 29, NO. 6, JUNE 1957

973

can be calculated for any compound \vhich will predict its backscattering with high precision. This quantity is defined as follows:

/c5c‘ (3

5 a

c W

where AB and AG are the atomic weights of B and C for the compound B,C,. So defined, Z is simply the sum of each atomic number multiplied by the weight fraction of that atom which is present in the compound. On this basis the value of Z calculated for calcite (CaCOJ is 12.565, which indicates that backscattering should be intermediate between magnesium (2 = 12) and aluminum (2 = 13). When the comparison is made with these substances, calcite yields a value of 2 within 0.08% of the calculated value. A more striking example may be chosen. Such diverse substances as halite (XaCI) and fluorite (CaF2) have the following 2 values : For CaF2, Z = 14 646 For NaCl, Z = 14 639 These differ by less than 0.05% and indicate that both substances should scatter intermediate between silicon (Z = 14) and phosphorus (2 = 15). Such is the case. [In practice it is more convenient t o interpolate between silicon (2 = 14) and sulfur (2 = 16).] It should not be inferred that the Z concept requires direct interpolation between the nearest elements, although this is often convenient. I n the above example, the 2 values for fluorite and halite can be substituted directly in the equation given in Table I for period I11 (2= 10 to 18). ‘Then this is done, the predicted values can he compared with direct measurement. yo Backscattering Calcd. Obsd. A CaF2 14.643 14.652 0.009 NaCl 14.636 14.628 0.008

If the backscattering for these substances is calculated from the intrinsic scattering of the elements (Table I) and their weight fraction, one gets CaFz = 15.944y0 and NaCl = 14.636%. This illustrates that it is immaterial which mode of computation is used when both atoms are in the same period. Another example may be cited to illustrate the general validity of the 2 concept. I n the case of zinc sulfide one gets the value, .2? = 25.393 (calcd.) which indicates that its backscattering should lie between manganese (25) and iron (26). Direct intercomparison with these metals yielded a value of Z = 25 368 (obsd.) which differs by less than 0.1% from the calculated value. If the observed Z value is substituted in the equation 974

ANALYTICAL CHEMISTRY

,I

/

a

==“f-01

0 2

0.3

04

WEIGHT

Figure 5. solutions

06

05

FRACTION

0 7

IO

09

08

OF S O L U T E

Backscattering of beta particles by alkali halide

Validity of extrapolation to unit weight fraction oi solute coilfirmed for XaC1. KCI, and RbCl by synthetic single crystal